OpenAI Agent SDK using local and Azure deployed LLMs

Introduction

This is a repository which shows how to use the Agent SDK with models local models or models deployed to Azure. The Agent SDK requires a model which support structured output. For more info :

• Details for Azure
• Details for Ollama

Setup

The repository is configured with a VsCode devcontainer with all necessary dependencies installed, and Poetry will install necessary Python dependencies.

Prerequisites

• Docker
• [Azure] Model deployed to Azure that supports tool usage.
• [Ollama] Ollama installed and a model running on your computer that supports tool usage.
  • How to install, run and pull models using Ollama.

Tasks

1. Make sure you select the poetry python environment prefixed with openai-agent.
2. Create a .env file in the root-folder based on the structure shown in .env_example

Known challenges

• The Agent SDK assume the model is named 'gpt-4o' if not otherwise instructed.